Abstract
Approximately 75% of coagulase-negative staphylococci are resistant to methicillin, but it is suspected that even more resistance exists that is not detected by standard susceptibility assays. To determine the most accurate assay for measuring resistance, we compared the detection of mecA by PCR with detection by National Committee for Clinical Laboratory Standards methods using oxacillin as the class drug. Strains from 11 species of coagulase-negative staphylococci were selected such that 84% were susceptible by the broth microdilution method. Of 45 mecA-positive strains, 1 strain was unable to express the mecA gene product after induction and was not included in further analyses. For microdilution with 2% NaCl, the disk test without salt, and agar screen containing 4% NaCl plus-6 micrograms of oxacillin per ml, the sensitivities in detecting the 44 mecA-positive strains were 50, 84, and 70%, respectively, at 24 h and 77, 82, and 100%, respectively, at 48 h. The specificities of microdilution, disk, and agar screen in detecting the 97 strains lacking mecA were 100, 89, and 100%, respectively, at 24 h. Only the disk test proved to be less specific at 48 h (81%). Furthermore, for 10 of the mecA-positive strains plus an additional 8 strains subsequently added to the analyses, the MICs were 2 micrograms/ml at 24 h by the broth microdilution method; all 18 strains were positive for mecA by PCR. Thus, an oxacillin MIC of > or = 2 micrograms/ml indicated resistance and is probably a more appropriate breakpoint than the current National Committee for Clinical Laboratory Standards breakpoint of 4 micrograms/ml for coagulase-negative staphylococci. Strains for which MICs are < 2 micrograms/ml may be methicillin resistant and should be verified as susceptible by oxacillin agar screening with incubation for 48 h.